R_cyph Posted October 18, 2012 Share Posted October 18, 2012 Hey all, I have a particle system that I'm using to drive the velocity of another particle system using Particle cloud nodes in Vops. I'm just wondering if there is a way to create a divergence free velocity field from the first system and use that to advect the 2nd system with the field instead I've searched high and low and the only info I can find about divergence free fields is about the curl noise node. Quote Link to comment Share on other sites More sharing options...
Macha Posted October 18, 2012 Share Posted October 18, 2012 Somehow bringing it into dops and calculating a projection there would be my first attempt. Dive in, take it apart and see if the projection is hackable? Quote Link to comment Share on other sites More sharing options...
R_cyph Posted October 18, 2012 Author Share Posted October 18, 2012 Hey, what am I taking apart, the curl noise? I had a look inside and i can't see exactly what makes it divergence free. I can do this with 3ds max and ember ( a plugin from thinkbox, makers of krakatoa ). , But it is done in an automated way and I have no idea whats going under the hood. Quote Link to comment Share on other sites More sharing options...
Macha Posted October 18, 2012 Share Posted October 18, 2012 I think out-of-the-box curl is the only divergence free noise. How about you do some random sims of stuff in a box -a la brownian motion- and then write out the vel field as a noise you can look up. That would be div free. Quote Link to comment Share on other sites More sharing options...
ikarus Posted October 19, 2012 Share Posted October 19, 2012 If you want a divergence-free velocity field based off a particle set I would start with the FLIP solver as it contains all the microsolvers involved in translating a particle set to a field set and simulating accordingly. Quote Link to comment Share on other sites More sharing options...
R_cyph Posted October 19, 2012 Author Share Posted October 19, 2012 Hey, I'm just watching the "fluid solver from the scatch master class" on the sidefx site. They go over transferring particle velocity to a volume vel field for anyone interested. Quote Link to comment Share on other sites More sharing options...
Aliencorpse Posted November 6, 2012 Share Posted November 6, 2012 The pressure projection is what makes a velocity divergence free, they are the gas project non divergence. Make an empty field, put your noise into it and then run the pressure projection on it. I think that will do it. Then you can use the resulting volume to advect shit through. Quote Link to comment Share on other sites More sharing options...
R_cyph Posted February 28, 2013 Author Share Posted February 28, 2013 I still haven't succesfully been able to do this. I've found some tutorials that go about this in a more indirect way. Using the pyro solver to create a divergence free velocity/ density field from point or geometry normals, then using the advect by volume node. I suspect this is very ineffecient. I'm trying to duplicate this effect in houdini dops, but I don't have the knowledge to build this from scratch. I've watched the "building fluid solver from scratch flip videos" but still can't figured it out. is very simple and straight forward in 3ds max ember. ( 1 particle system converted to divergence free velocity field then used to advect another )pretty please help Quote Link to comment Share on other sites More sharing options...
Filipp Elizarov Posted February 28, 2013 Share Posted February 28, 2013 (edited) - I made a quick example for you.Maybe it will help you to achive your goal. 1.Create particles. 2.Cache them 3.Create dopnetwork and import particles to dopnet with support of sop solver 4.create empty vel vector field 5.create empty tempvel field (same as vel) which will contain V from particles 6.Gas Particles To Field - V to Temp vel 7.Gas linear combination add temp vel to vel 8.Gas project non divergent (for vel field) after go to sop level and load vel field from dop to sop and use it in POP through Advect by Volume. get_non_divergent_vel_from_particles_through_DOP.hip Edited February 28, 2013 by Filipp Elizarov Quote Link to comment Share on other sites More sharing options...
R_cyph Posted February 28, 2013 Author Share Posted February 28, 2013 Geez. I've been trying to get this working on and off for months lol. May I ask how you figured this out? have you been using houdini for forever? Thanks for you help! Quote Link to comment Share on other sites More sharing options...
R_cyph Posted February 28, 2013 Author Share Posted February 28, 2013 (edited) quick question in the "gaslinearcombination" node . Why are you adding the vel to the tempvel field? the velocity of the particle remains in the field after it has moved on. I tried setting the operation to maximum or average which seems to let the velocities dissipate or be overwritten instead of just adding on top of eachother. Do we have to add the tempvel to vel at all? Any insight into this would be greatly appreciated. Edited February 28, 2013 by R_cyph Quote Link to comment Share on other sites More sharing options...
Filipp Elizarov Posted February 28, 2013 Share Posted February 28, 2013 HI here is nothing strange - vel in current step = vel from previous step + temp vel. (vel = vel+tempvel).I use Houdini everyday at work =) Maybe the book Fluid simulation for computer graphics will help you understand it a bit better and practice practice practice from very simple scenes to more complicated. Quote Link to comment Share on other sites More sharing options...
R_cyph Posted February 28, 2013 Author Share Posted February 28, 2013 If I set the gasparticletofield to vel and turn off the gaslinearcombinatio node I get the result I'm looking for. Since the particles velocity is already cached I don't need to add look at previous timestep of the velocity field back in, right? Quote Link to comment Share on other sites More sharing options...
Netvudu Posted March 2, 2013 Share Posted March 2, 2013 just curious...what´s the usefulness in getting this divergence-free velocity field? I suppose you won´t be going through this without a strong reason... Quote Link to comment Share on other sites More sharing options...
edward Posted March 2, 2013 Share Posted March 2, 2013 just curious...what´s the usefulness in getting this divergence-free velocity field? http://en.wikipedia.org/wiki/Incompressible_flow Quote Link to comment Share on other sites More sharing options...
Netvudu Posted March 2, 2013 Share Posted March 2, 2013 from that wikipedia article: And so beginning with the conservation of mass and the constraint that the density within a moving volume of fluid remains constant, it has been shown that an equivalent condition required for incompressible flow is that the divergence of the fluid velocity vanishes. So the goal of having a divergence free field is an incompressible fluid? That´s cool, but I thought Flip fluids were already incompressible fluids. Why creating those this way? Quote Link to comment Share on other sites More sharing options...
Filipp Elizarov Posted March 2, 2013 Share Posted March 2, 2013 (edited) Sure it isn't a good idea to do a lot of work which you don't need )But maybe the person who ask about it - really need it. Edited March 2, 2013 by Filipp Elizarov Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.